The Problem of Rarity: Estimation of Prevalence in Rare Disease.

Background: From a disease's first description to its wider recognition, factors such as changes over time in diagnostic criteria, available therapies, and subsequent mortality rates may influence diagnosed prevalence of rare diseases.

Objectives: To propose a novel methodology for estimating the true prevalence of rare diseases using current incidence adjusted to changing diagnostic practice over time. This article focuses on rare diseases whose diagnosis may have changed over time, and raises the hypothesis that prevalence calculated from current incidence may be higher than diagnosed prevalence, which may lag behind the current disease definition and diagnostic methods. A rare epileptic encephalopathy, Dravet syndrome (DS), is explored as an illustrative example.

Methods: A targeted literature review was performed for DS to identify all reported incidence, prevalence, and mortality and depict how diagnostic practice has evolved over time. A conceptual model was developed to calculate prevalence derived from current incidence figures alone (incidence-derived prevalence) or incidence adjusted with factors that cause a diagnostic drag (diagnostic awareness-adjusted prevalence).

Results: We identified sufficient publications of incidence and prevalence to test the conceptual model. For pediatric patients with DS, diagnosed prevalence in the field (as reported in current literature) matches incidence-derived prevalence, whereas for adult patients, it is overestimated by incidence-derived prevalence, but not by diagnostic awareness-adjusted prevalence.

Conclusions: Care should be taken with current incidence-derived prevalence figures to not overstate the prevalence in rare diseases, as methodological challenges in counting small populations, coupled with advances in rare disease discovery, may cause discrepancies.